By John Edwards
NOTE: This is an overview of the entire article, which appeared in the May 2012 issue of the IEEE Signal Processing Magazine.
Click here to read the entire article.
In labs around the world, researchers are working on a new generation of wireless medical sensors and related networking technologies. Wearable sensors, whether functioning alone or connected into multinode body areas networks (BANs), are designed to help clinicians and other caregivers accurately monitor patients’ temperature, heart rate, brain activity, muscle motion, and other critical data.
Sensor technology developers as well as many medical industry business analysts feel that standalone sensors and BANs have the potential to both improve health care and lower costs by nipping lurking medical problems in the bud and speeding rehabilitation. However, it remains to be seen how large the impact will be.
The article cites three examples of how wireless sensor research is being pursued.
IMEC/Holst Centre, a wireless solutions research institute located in Eindhoven, The Netherlands, for instance, is addressing multiple conditions, including epilepsy, heart disease, and sleep apnea, with its sensor-driven prototype body monitoring system.
BANs, such as the kind being developed at IMEC/Holst Centre, are designed to provide either continuous or time-frame-delineated insight into potentially life-threatening conditions.
The technology can also be used on post-operative patients. “If they want to send you home but still want to check up on you,” says Harmke De Groot, ultralow-power wireless and digital signal processing (DSP) program director at IMEC/Holst Centre. Other potential users include people living in nursing homes and other types of long-term care facilities.
At the University of Bologna, Dr. Lorenzo Chiari, an assistant professor of biomedical engineering, is developing sensors for the European Union-sponsored project called CuPid, which stands for “closed-loop system for personalized and at-home rehabilitation of people with Parkinson’s disease,” a project that aims to provide personalized rehabilitation exercises for home-bound people with Parkinson’s disease (PD).
Sensor-delivered audio biofeedback can help a patient with PD maintain control over body movements.
To address the needs of PD patients, CuPiD partners are developing a home-based personalized rehabilitation system including wearable sensors that enable audio biofeedback, virtual reality images, and other rehabilitative services. Wireless sensor technology is crucial for meeting CuPid’s objective of creating a system that doesn’t physically hinder people who already face significant mobility challenges.
Researchers are also hoping to use wireless sensor technology to create around-the-clock patient monitoring environments for hospitals and other medical facilities. Barnes-Jewish Hospital in St. Louis, Missouri recently tested a sensor network that allows vital signs to be tracked as patients move about the facility during convalescence or while waiting for tests or other procedures. The system, developed by researchers at nearby Washington University, monitored at-risk patients, measuring blood oxygenation and heart-rate readings at rates of once or twice a minute. The data was then transmitted to a base station where it was merged with other information in the patient’s electronic medical record, such as lab test results. Data collected by the system was continuously scrutinized by a machine-learning algorithm for any indications of clinical deterioration. Whenever a potential problem was detected in a patient’s data stream, the system automatically placed a phone call to a nurse or other designated caregiver, alerting the individual to check on the patient. The idea behind the test system was to create a virtual intensive care unit (ICU) where the patients aren’t wired to beeping machines and instead are free to move about as they please.
The overall reliability of such systems is an issue, of course. The article addresses the technology used to increase overall reliability, and provides an early measure of deployed reliability.
The biggest barriers to deploying these types of systems is power and device size. For now, IMEC/Holst Centre and Eindhoven-based semiconductor manufacturer NXP have developed an ultralow power biomedical signal processor, the CoolBio, that’s designed to meet the energy requirements of planned BANs. De Groot notes that the CoolBio allows BAN sensor nodes to draw less power. Processing and data compression is performed locally on the BAN node, limiting power-wasting wireless data feeds while simultaneously adding motion artifact reduction and smart diagnosis functions.
Although these challenges are still not entirely conquered, researchers feel that medical sensors are destined to play a major role in improving patient care and cutting costs. Dr. Chengyan Lu, of Washington University states, “This is one area where engineering and science can really make a revolutionary contribution to society.”